JP2011504324A - Synchronized multilink transmission in an ARQ-enabled multihop wireless network - Google Patents

Abstract

An excellent method for synchronized multi-link transmission in an ARQ-compatible multi-hop wireless network is provided.
A method and apparatus for synchronized multilink transmission in an ARQ-enabled multihop wireless network is disclosed herein. In one embodiment, the method performs pre-transmission of packets to multiple hops, and multiple hops in a base station and network transmit packets synchronously to one or more mobile stations in a wireless communication system. One or more times of the packet if the remaining time to the synchronized transmission time for the entire path is greater than a threshold at the set of steps and one or more hops forming the path in the network Performing the retransmission.
[Selection] Figure 2

Description

  The present invention relates generally to multi-hop wireless networks and radio stations in such networks, and more particularly to transmission in ARQ-enabled multi-hop networks that support synchronized transmission over multiple links.

(priority)
This patent application is a copy of the corresponding US Provisional Patent Application No. 60 / 985,162 entitled “Synchronized Multi-Link Transmission in an ARQ-Enabled Multi-Hop Wireless Network” filed on Nov. 2, 2007. Priority is claimed and 60 / 985,162 is incorporated by reference.

(Related application)
This application is assigned to the corporate assignee of the present invention and is filed on September 16, 2008, entitled “A METHOD FOR ARQ-ENABLED PRE-TRANSMISSIONS IN MULTICAST AND BROADCAST SERVICES OF RELIZ NETWORKS UTILIDS Related copending applications, U.S. patent application Ser. No. 12 / 211,611.

  FIG. 1 illustrates an exemplary multi-hop wireless network. Referring to FIG. 1, depending on how a station functions for transmission of packets to / from an external network, a BS (base station), RS (relay station), and SS ( A number of radio stations that can be classified as subscriber stations). An SS (subscriber station) such as SS 101 in FIG. 1 may or may not be mobile. When a subscriber station is mobile, it is sometimes represented as an MS (Mobile Station).

  In the downlink, the base station 102 introduces external network packets into the multi-hop network, while the subscriber station 101 consumes or terminates these packets. In the uplink, the subscriber station originates packets while the base station 102 introduces these packets to the external network. A relay station such as RS1, RS2, or RS3 exists on the path between the base station 102 and the subscriber station 101 and relays packets for the uplink or downlink. As used herein, subscriber stations can include, but are not limited to, telephones, laptops, desktops, PDAs, and the like.

  Through one or more switches and / or routers (not shown), the base station 101 is connected to an external network (not shown), which can be a wired network or a wireless network. The external network can again be a multi-hop wireless network.

  In a multi-hop wireless network, a radio link can serve as an access link, a relay link, or both an access link and a relay link. Referring to FIG. 1, a wireless link acts as an access link when a packet transmitted over the link is received or transmitted by a subscriber station connected to the link. A radio link acts as a relay link when packets transmitted via this link are received and transmitted by a base station or a relay station. A radio link serves as both a relay link and an access link when packets transmitted over this link are received by both the relay station and the subscriber station.

  The coverage of the base station can take various shapes depending on antennas, obstacles, channel conditions, and the like. Depending on the location and deployment of base stations and relay stations, a subscriber station may be within the coverage of one or more other radio stations.

  To improve reception performance using macro diversity, transmissions over multiple links are often synchronized. For example, transmissions on multiple access links associated with a subscriber station can be synchronized to improve the reception performance of the subscriber station. The transmission can be a unicast transmission to a specific station (as in soft handoff and coordinated communication), or it can be a multicast or broadcast transmission to multiple stations (multicast service and As in broadcast service). As another example, transmissions on multiple relay links associated with the relay station or the base station can be synchronized to improve the reception performance of the relay station or base station. Transmissions on synchronized multiple links can be uplink (in the direction towards the base station) or downlink (in the direction towards the subscriber station). Hereinafter, related techniques and embodiments of the present invention are described herein with respect to synchronized transmissions on the access link, which is the downlink. Thus, in general, a synchronized link can be an access link, a relay link, etc., but is assumed to be an access link. Also, in general, packets related to synchronized transmissions can be generated or consumed by any of the base station, relay station, and subscriber station, and these packets can also be transmitted in the downlink direction or It is assumed that although it can be transmitted in the uplink direction, it is introduced by the base station, distributed in the downlink direction, and consumed by the subscriber station. However, it should be noted that the present invention can be easily extended to the general case where such assumptions may not necessarily hold.

  Upon receiving multiple signals from multiple transmitters (each transmitter is a base station or relay station), the subscriber station applies diversity combinations, such as a selection combination and a maximum ratio combination, to generate an improved signal. Obtainable.

  Synchronized transmission requires delay control on the path to the access link where the synchronized transmission is to take place. Due to the difference in the number of relay stations in the path, the difference in propagation delay in each link, the difference in processing delay in each relay station, and the difference in channel status, delays on the path may not necessarily be the same. Thus, the transmission time for synchronized transmission is generally determined by taking into account the longest delay among the delays in all paths. For packet transmission on a path other than the one that suffers the longest delay, an artificial delay must be introduced so that transmissions on multiple links are synchronized. Synchronized transmission methods for multicast and broadcast services are known. In one such method, each relay station reports its processing delay to the base station as a capability parameter. The base station then calculates the maximum cumulative delay of all relay stations in the multicast and broadcast coverage based on the position of these relay stations in the tree and their individual processing delays. To do. Next, the base station calculates the required waiting time Wi for each relay station based on the value of the maximum accumulated delay and the accumulated delay of each relay station, and sends the waiting time of the relay station to each relay station. To be notified. If synchronized multicast and broadcast transmissions are required, the base station forwards the data on the relay downlink by a number of frames equal to the maximum accumulated delay as a pre-transmission and then over this access link Transmit data. Each relay station in the multicast and broadcast area forwards data that the relay station receives on the relay downlink. Finally, after the base station waits for a number of frames equal to the maximum accumulated delay, and after each relay station waits for the relay station's designated waiting time, Wi, the base station and the relay station pass over the access link. Data is transmitted synchronously. However, this particular approach does not consider any channel errors in the radio link. If a channel error occurs for a packet transmitted on a link, all relay stations and access stations in the forward path from that link will not be able to receive the packet correctly, so the packet will be synchronized. Cannot be used for transmitted transmissions.

  In another synchronized transmission method for multicast and broadcast services, each relay station reports its processing delay as a capability parameter to the base station, and then the base station in the multicast and broadcast service area The maximum cumulative delay of all relay stations is calculated based on the position of these relay stations in the tree and the individual processing delays of these relay stations, as in the case of the synchronized transmission method described above. In contrast, in this synchronized transmission method, the base station determines a target transmission frame on the access link for each synchronized data burst based on the maximum accumulated delay and other base station information. . When ARQ is enabled on the relay link, the base station can select the target transmission frame to accommodate the delay due to ARQ retransmission on the relay link. The base station includes the frame number of the target transmission frame in each packet. The relay station transmits the packet to the subscriber station via the access link in the target transmission frame. The intermediate relay station relays the packet to a relay station below the intermediate relay station based on the restriction of the QoS parameter regarding the relay connection regarding the packet. During a connection setup for a subscriber station, if a relay path with appropriate characteristics, such as a QoS configuration per hop, is not available, the base station will create a new relay with the required characteristics, such as a QoS configuration per hop. You can start creating a path. Any QoS parameter of the configured hop-by-hop connection on the existing relay path can be changed by the base station.

  Also, in this synchronized transmission method, if the relay station cannot transmit a packet in the target transmission frame of the packet, the relay station informs the base station of the failure. The relay station includes the delay time of arrival of the packet in units of frames. Further, if the relay station determines that the packet arrived earlier than the relay data early arrival report threshold for the target transmission frame of the packet, the relay station can supply early arrival information to the base station. Upon detecting an early arrival, the relay station includes the number of frames that exceeded the threshold, waiting for the packet to be transmitted. If the relay data early arrival report threshold is set to 0, early arrival reporting is disabled.

  FIG. 2 shows an exemplary ARQ state diagram for the transmitter. When the packet is transmitted, the ARQ state transitions to an “unresolved” state. When an ACK is received in ARQ_RETRY_TIMEOUT, the ARQ state transitions to the “complete” state. If ARQ_RETRY_TIMEOUT expires without receiving an ACK, or if a NACK is received from the receiver, the packet is retransmitted while the ARQ state changes from “unresolved” to “waiting for retransmission”. Next, transition from “waiting for retransmission” to “unresolved”. If ARQ_BLOCK_LIFETIME expires while the ARQ state is in the “unresolved” state or “waiting for retransmission” state, the packet is discarded. Thus, the packet can be correctly transmitted or discarded after ARQ_BLOCK_LIFETIME, and the longest delay occurs when the packet is discarded. ARQ_BLOCK_LIFET IME is selected such that one or more retransmissions can occur before the lifetime expires.

  Disclosed herein is a method and apparatus for synchronized multilink transmission in an ARQ-enabled multihop wireless network. In one embodiment, the method performs pre-transmission of packets to multiple hops, and multiple hops in a base station and network transmit packets synchronously to one or more mobile stations in a wireless communication system. One or more times of the packet if the remaining time to the synchronized transmission time for the entire path is greater than a threshold at the set of steps and one or more hops forming the path in the network Performing the retransmission.

  The invention will be more fully understood from the detailed description given below and from the accompanying drawings of various embodiments of the invention, the detailed description and the accompanying drawings, in which: It should not be construed as limited to the particular embodiments, but is for explanation and understanding only.

It is a figure which shows a multihop network. FIG. 4 is an ARQ state diagram for a transmitter. 6 is a flow diagram illustrating one embodiment of a retransmission process. 6 is a flow diagram illustrating one embodiment of a scheduling process. It is a figure which shows an example topology. It is a figure which shows another topology example. 5 is a flow diagram illustrating one embodiment of a process for handling packets performed by a base station. It is a figure which shows one Embodiment of a frame format. 2 is a flow diagram for one embodiment of a process for performing ARQ at a transmitter. FIG. 3 illustrates one embodiment of an ARQ state diagram for a transmitter. 6 is a flow diagram illustrating one embodiment of a process for handling packets performed by a relay station. 2 is a flow diagram illustrating one embodiment of a process for determining whether a packet is in time for a synchronized transmission. FIG. 2 is a block diagram illustrating an embodiment of a station.

  A method and apparatus for enabling synchronized multilink transmission in an ARQ-enabled multihop network without incurring unnecessarily large delays is described.

  In the following description, numerous details are set forth to provide a more thorough explanation of the present invention. However, it will be apparent to those skilled in the art that the present invention may be practiced without these specific details. In other instances, well-known structures and devices are shown in block diagram form, rather than in detail, in order to avoid obscuring the present invention.

  Some portions of the detailed descriptions that follow are presented in terms of algorithms and symbolic representations of operations on data bits within a computer memory. These algorithmic descriptions and representations are the means used by data processing technology vendors to most effectively communicate their work to other peers. The algorithm is, in this case, and generally also considered a self-consistent sequence of steps leading to the desired result. These steps are those requiring physical manipulation of physical quantities. Usually, though not necessarily, these quantities take the form of electrical or magnetic signals capable of being stored, transferred, combined, compared, and otherwise manipulated. Occasionally, it has proven convenient to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, etc., mainly for general usage reasons.

  It should be noted, however, that all of these and similar terms are to be associated with the appropriate physical quantities and are merely convenient labels attached to these quantities. Unless stated otherwise, as will be clear from the description below, throughout this description, such as “processing” or “calculating” or “calculating” or “determining” or “displaying”, etc. An explanation utilizing terminology manipulates data represented as physical (electrical) quantities in computer system registers and in memory, in computer system memory, or in computer system registers, or other such It is understood to refer to actions and processes of a computer system or similar electronic computing device that translates into other data that is also represented as a physical quantity within an information storage device, information transmission device, or information display device. .

  The invention also relates to an apparatus for performing the operations herein. The apparatus can be specially constructed for the required purposes, or comprises a general purpose computer that is selectively activated or reconfigured by a computer program stored in the computer. It is possible. Such a computer program can be any type of disk, including a flexible disk, optical disk, CD-ROM, and magneto-optical disk, ROM (read only memory), RAM (random access memory), EPROM, EEPROM, magnetic card or optical Stored in a computer readable storage medium coupled to a computer system bus, such as, but not limited to, a card or any type of medium suitable for storing electronic instructions Is possible. The algorithms and displays presented herein are not inherently related to any particular computer or other apparatus. Various general purpose systems can be used with programs according to the teachings herein, or it may prove advantageous to build a more specialized device that performs the required method steps. There is sex. The required structure for a variety of these systems will appear from the description below. In addition, the present invention is not described with reference to any particular programming language. It will be appreciated that a variety of programming languages can be used to implement the teachings of the invention as described herein.

  A machine-readable medium includes any mechanism for storing or transmitting information in a form readable by a machine (eg, a computer). For example, machine-readable media include read only memory (“ROM”), random access memory (“RAM”), magnetic disk storage media, optical storage media, flash memory devices, electrical form, optical form, acoustic form, or others And propagated signals (eg, carrier waves, infrared signals, digital signals, etc.).

(Outline)
An apparatus and method for transmission in an ARQ (automatic repeat request) capable multi-hop network that supports synchronized transmission over a specific set of links (referred to herein as synchronized links) is described. . FIG. 3A is a flow diagram of one embodiment of a retransmission process. This process is performed by processing logic that can comprise hardware (circuitry, dedicated logic, etc.), software (such as is run on a general purpose computer system or a dedicated machine), or a combination of hardware and software. Is done.

  Referring to FIG. 3A, the process is such that processing logic performs pre-transmission of packets to hops so that the base station and those multiple hops in the network are one or more mobile stations in the wireless communication system. Begin by enabling the packets to be transmitted synchronously (processing block 311). Next, processing logic determines the packet's 1 if the time remaining in the set of one or more hops forming the path in the network before the synchronized transmission time for that entire path is greater than the threshold. One or more retransmissions are performed (processing block 312). In one embodiment, this threshold is a QoS requirement that includes link retransmission delay for links on the forward path, channel characteristics of links on the forward path, multi-hop network topology, delay requirements and / or packet loss requirements. , Based on one or more of the group consisting of the normal forward path delay and the path retransmission budget for that path. In another embodiment, this threshold is zero. In yet another embodiment, the threshold is a normal forward normal path delay. In yet another embodiment, the threshold is the sum of normal forward path delay and one or more link retransmission delays for one or more links in the forward path. In a further embodiment, this threshold is the sum of normal forward path delay and ARQ block lifetime for one or more links in the forward path.

  In one embodiment, the method calculates a synchronized transmission time of a packet over a synchronized link, and in such cases, in a set of one or more hops that form a path in the network, If the time remaining until the synchronized transmission time for the entire path is greater than the threshold, between any two stations on the path according to ARQ on at least one of the links on the path It further includes performing one or more retransmissions of the packet by retransmitting the packet.

  In one embodiment, the synchronized transmission time of the packet over the synchronized link is calculated based on the transmission time offset and the link transmission time of the packet. In such cases, in one embodiment, the method calculates the normal link delay and link retransmission delay for each relay link on each path from the base station to the access station associated with the synchronized link. And calculating a transmission time offset based on a normal link delay and a link retransmission delay, and calculating a link transmission time of a packet via the next relay link when a packet arrives at the base station. In one embodiment, the transmission time offset is calculated based on taking a maximum value between the path delay budgets for all paths. In one embodiment, when calculating the transmission time offset, the path delay budget for each path from the base station to the access station is calculated based on the sum of the normal path delay and the path retransmission budget. In one embodiment, the link transmission time of a packet over the next relay link is based on one or more of packet arrival time, base station packet processing time, and packet QoS (Quality of Service) provisioning. In another embodiment, the link retransmission delay for a link is based on the time required for ARQ retransmission over that link. In yet another embodiment, the link transmission time of a packet over the next relay link is calculated when the packet arrives at the base station.

  In one embodiment, when calculating the transmission time offset, the normal path delay for each path from the base station to the access station is based on the sum of all normal link delays for the links on that path. Is calculated. In another embodiment, the normal forward path delay for each radio station on the path from the base station to the access station is calculated based on the sum of the normal link delays for the links from the radio station and the access station. .

  In one embodiment, the path retransmission budget for each path from the base station to the access station is calculated based on the link retransmission delay for the link on that path when calculating the transmission time offset. In another embodiment, the path retransmission budget may include channel characteristics of links on the path, multi-hop network topology, QoS requirements including delay requirements and / or packet loss requirements, normal path delays, and path retransmissions for other paths. Based on one or more of a group of budgets. In yet another embodiment, the path retransmission budget for the path is large enough to accommodate the link retransmission delay for the link on the path from the base station to the access station, but less than the sum of the link retransmission delay for that path. .

  In one embodiment, the process of FIG. 3A further includes transmitting the packet at the synchronized transmission time of the packet.

  In one embodiment, the process of FIG. 3A further includes discarding the packet on the ARQ-enabled link. This can be done by the transmitting station sending an ARQ discard message for the packet to the receiving station if the time remaining until the synchronized transmission time for the entire path is less than the threshold. There is sex. In one embodiment, discarding a packet means that upon receiving an ARQ discard message for the packet from the transmitting station, the receiving station discards the packet and sends an acknowledgment for the ARQ discard message. Including. In another embodiment, discarding the packet includes the station sending forward the ARQ window associated with the packet. In yet another embodiment, discarding a packet on the link expires the ARQ block lifetime for that packet, and further the path retransmission budget for the path is linked to the link retransmission delay of the link on the path from the base station to the access station. This is done if it is large enough to accommodate but less than the sum of the ARQ block lifetimes for all links in the path. In yet another embodiment, discarding a packet on a link means that the number of packet retransmissions exceeds the retry limit for that link, and the path retransmission budget for the path is on the path from the base station to the access station. This is done if it is large enough to accommodate the link retransmission delay of the link, but less than the total time required for the retransmission retry limit.

  FIG. 3B is a flow diagram of one embodiment of a process for calculating various delay parameters that may be used for synchronized multilink transmission. This process is performed by processing logic that can comprise hardware (circuitry, dedicated logic, etc.), software (such as is run on a general purpose computer system or a dedicated machine), or a combination of hardware and software. Is done. This process can be performed in a centralized manner or in a distributed manner. In one embodiment, a centralized scheduler at the base station performs this entire process. In another embodiment, the scheduler is distributed across multiple radio stations, and some or all of the operations in FIG. 3B are performed in a distributed manner by base stations, relay stations, or subscriber stations. In such cases, the results of operations performed by the wireless station are exchanged with other wireless stations whenever necessary.

  Hereinafter, for ease of understanding, embodiments of the present invention will be described in the context of a centralized scheduler, unless otherwise specified. Reference is also made to the multi-hop network shown in FIG.

  Referring to FIG. 3B, the scheduler determines a network topology for multilink synchronized transmission (processing block 301). This topology can be the same as the topology of the multi-hop network. For example, if the multilink synchronized transmission is a broadcast transmission, the base station and all relay stations in the multihop network will participate in the multilink synchronized transmission over the access link and thus their The topology is the same. On the other hand, this topology is multi-hop depending on the location of subscriber stations, the coverage of base stations and relay stations, the type of transmission (whether the transmission is unicast, multicast or broadcast), etc. It may be a subset of the topology of the network.

  FIG. 4 shows an exemplary network topology constructed from the multi-hop network shown in FIG. In this example, there is only one subscriber station 402 that receives synchronized multilink transmissions, and the subscriber station 402 is within the coverage of two relay stations, RS1 and RS2. There are two access links to the subscriber station 402: a relay station, one link from RS1, and the other link from the relay station, RS2. There are two relay paths leading to the access link, and each path includes one relay link.

  FIG. 5 shows another exemplary network topology constructed from the multi-hop network shown in FIG. Unlike the previous example of FIG. 4, the subscriber station 502 in this example is within the coverage of three relay stations, RS1, RS2, and RS3. There are three access links, each from one of the three relay stations to the subscriber station 502. There are also three relay paths from the base station 501 to the access link. Thereafter, these relay paths and links on the paths are indexed as follows for the sake of explanation. The relay paths from the base station 501 to the relay stations RS1, RS2, and RS3 are indexed with the first path, the second path, and the third path, respectively. Similarly, the links for each path are indexed starting from the link connected to base station 501. The first path comprises only one link from the base station 501 and the relay station RS1, and this link is indexed with the first link of the first path. The second path again comprises only one link from the base station 501 and the relay station RS2, and this link can be similarly indexed. On the other hand, the third path comprises two links, one link leading from the base station 501 to the relay station, RS2 (indexed with the first link of the third path), and the other link is From the relay station, RS2, to the relay station, RS3 (indexed with the second link of the third path).

  Note that the second path (the relay path from the base station 501 to the relay station RS2) is a sub-path of the third path (the relay path from the base station 501 to the relay station RS3). As far as FIG. 5 is concerned, there is no need to consider a relay path in a subpath of another relay path, and the resulting delay parameter does not change with or without such consideration. Thus, for the scheduling example, only two relay paths are considered.

  Referring again to FIG. 3B, processing logic calculates a per-link delay parameter for each link on the relay path (processing block 302). In one embodiment, the per-link delay parameter includes two different parameter sets. One set of delay parameters relates to the delay experienced by the initial transmission of a packet, regardless of whether ARQ is enabled on the link. In one embodiment, this delay parameter set can be used for packet processing delays such as address matching, packet classification, and routing, scheduling delays, link propagation delays, packet transmission delays, and packet routing delays when the relay performs a particular process. It includes other processing delays required for processing and transmission, but does not include the artificial delay provided to synchronize multilink transmissions. The other delay parameter set is related to the delay required for packet retransmission when ARQ is enabled. In one embodiment, this delay parameter set includes a delay for the receiver to send an ACK or NACK, a delay for the transmitter to wait before initiating a retransmission (eg, ARQ_RETRY_TIMEOUT in FIG. 2), and the packet to Including the delay required to do. If ARQ is not enabled on the link, these parameter sets may be unnecessary or set to appropriate default values.

が生成される。同様に、パスｉの各リンクｊに関して、再送と関係する遅延が組み合わされて、リンク上で一巡のＡＲＱパケット再送に要求される時間を示す、リンク再送遅延

が生成される。一実施形態において、パスｉのリンクｊに関してＡＲＱが使用不可にされる場合、リンク再送遅延

は、０に設定される。 In one embodiment, for each link j in path i, the normal link delay, combined with the delay associated with the initial transmission, indicates the time required for successful packet transmission.

Is generated. Similarly, for each link j in path i, a link retransmission delay indicating the time required for one round of ARQ packet retransmission on the link, combined with the delay associated with retransmission.

Is generated. In one embodiment, if ARQ is disabled for link j in path i, link retransmission delay

Is set to 0.

として表される１つのパラメータは、パス上のリンクのいずれにおいても再送なしに、基地局から、パスの他方の終端の中継局まで伝送されることに成功したパケットが経験する遅延と関係する。一実施形態において、パスの通常のパス遅延は、パス上のすべてのリンクの通常のリンク遅延の合計として算出される。例えば、第ｉ番のパスの通常のパス遅延は、

であり、ただし、Ｌ_ｉは、第ｉ番のパス上のすべてのリンクのセットである。 Next, processing logic calculates a delay parameter for each path for each relay path (processing block 303). Again, the delay parameter for each path includes at least two different parameters. The normal path delay of path i in this specification,

One parameter expressed as is related to the delay experienced by a packet successfully transmitted from the base station to the relay station at the other end of the path without retransmission on any of the links on the path. In one embodiment, the normal path delay of the path is calculated as the sum of the normal link delays of all links on the path. For example, the normal path delay of the i-th path is

Where L _i is the set of all links on the i th path.

として本明細書に表される他方の遅延パラメータは、パケットを再送するのに使用されることが可能な時間と関係する。再送のための合計時間が、パス再送予算を超えない限り、再送は、任意の回数、そのパスの任意のリンクで行われることが可能である。このことは、確保された再送時間が、いくつかの仕方で複数のリンクによって共有されることを可能にする。また、このことは、複数のリンクによる確保された再送時間の適切な共有を実施することによって、品質向上（再送を介する）と、パケット伝送の遅延との間でトレードオフを行うことができる柔軟性を与える。 Pass retransmission budget for pass i,

The other delay parameter expressed herein as being related to the time that can be used to retransmit a packet. As long as the total time for retransmission does not exceed the path retransmission budget, retransmission can be performed any number of times on any link on that path. This allows reserved retransmission times to be shared by multiple links in several ways. In addition, this is a flexible way to make a trade-off between quality improvement (via retransmission) and packet transmission delay by appropriately sharing the retransmission time secured by multiple links. Give sex.

は、以下のとおり、適切に加重されたリンク再送遅延を合計することによって算出される。すなわち、

ただし、Ｌ_ｉは、第ｉ番のパス上のすべてのリンクのセットである。 In one embodiment, the path retransmission budget for a path is calculated based on the link retransmission delay for links on that path. In one embodiment, a path retransmission budget for path i,

Is calculated by summing appropriately weighted link retransmission delays as follows: That is,

However, L _i is a set of all links on the i-th path.

ただし、ｋは、再送の回数であり、更にｐ_{ｉ，ｊ，ｋ}は、第ｉ番のパスの第ｊ番のリンク上のｋ回の再送の確率である。別の実施形態において、パスｉのパス再送予算、

は、図２のＡＲＱ＿ＢＬＯＣＫ＿ＬＩＦＥＴＩＭＥなどの、適切に加重されたＡＲＱパラメータを合計することによって、算出される。一実施形態において、パス再送予算は、以下のとおり、算出される。すなわち、

ただし、Ｌ_ｉは、第ｉ番のパス上のすべてのリンクのセットである。ＡＲＱ＿ＢＬＯＣＫ＿ＬＩＦＥＴＩＭＥ_ｉ，ｊは、送信機が、パケットを破棄するまでに、ＡＣＫを受信することなしに待つ可能性がある最大時間を示す、第ｉ番のパスの第ｊ番のリンクのＡＲＱパラメータである。一実施形態において、加重パラメータ、ｂ_ｉ，ｊは、１以下である負でない実数である。より具体的には、一実施形態において、ｂ_ｉ，ｊは、送信機のＡＲＱタイマが、ＡＣＫを受信することなしにＡＲＱ＿ＢＬＯＣＫ＿ＬＩＦＥＴＩＭＥ_ｉ，ｊに達する確率によって決まる。 The weighting parameter, a _{i, j,} is a non-negative real number indicating the number of retransmissions on the j-th link of the i-th path. This number is not necessarily an integer, and this number can also be calculated probabilistically. For example, a _{i, j} can be calculated as the average number of retransmissions considering the retransmission probability as follows. That is,

Here, k is the number of retransmissions, and p _{i, j, k} is the probability of k retransmissions on the j-th link of the i-th path. In another embodiment, a path retransmission budget for path i,

Is calculated by summing appropriately weighted ARQ parameters, such as ARQ_BLOCK_LIFETIME in FIG. In one embodiment, the path retransmission budget is calculated as follows: That is,

However, L _i is a set of all links on the i-th path. ARQ_BLOCK_LIFETIME _{i, j} is the ARQ parameter of the jth link in the i th path indicating the maximum time that the transmitter can wait without receiving an ACK before discarding the packet . In one embodiment, the weighting parameter, b _{i, j} is a non-negative real number that is less than or equal to one. More specifically, in one embodiment, b _{i, j} is determined by the probability that the transmitter's ARQ timer will reach ARQ_BLOCK_LIFETIME _{i, j} without receiving an ACK.

は、パケットが、パス上の各リンクにおいて、ＡＲＱ機構によって許される最大回数で再送される最悪ケースシナリオに適応するのに十分なだけ大きい。いくつかのｊに関して、ａ_ｉ，ｊ＜ＲＥＴＲＹ＿ＬＩＭＩＴ_ｉ，ｊである場合、又はいくつかのｊに関して、ｂ_ｉ，ｊ＜１である場合、パス再送予算、

は、或るシナリオ（例えば、パス再送予算が、上流のリンクにおけるリンクによって使い果たされる）において要求される再送の回数に適応しない可能性がある。すべてのｊに関して、ａ_ｉ，ｊ＝０である場合、又はすべてのｊに関して、ｂ_ｉ，ｊ＝０である場合、パスｉ上で再送は全く行われ得ない。 For all j, if a _{i, j} = RETRY_LIMIT _{i, j} , or for all j, if b _{i, j} = 1, the path retransmission budget,

Is large enough to accommodate the worst case scenario where a packet is retransmitted at each link on the path the maximum number of times allowed by the ARQ mechanism. For some j, if a _{i, j} <RETRY_LIMIT _{i, j} , or for some j, if b _{i, j} <1, then the path retransmission budget,

May not adapt to the number of retransmissions required in certain scenarios (eg, the path retransmission budget is used up by links in upstream links). For all j, if a _{i, j} = 0, or for all j, b _{i, j} = 0, no retransmissions can occur on path i.

In one embodiment, the weighting parameters a _{i, j} and b _{i, j} include the following: QoS including link channel characteristics, multi-hop network topology, ARQ mechanism, delay requirements and / or packet loss requirements on the path. It is calculated taking into account one or more of the requirements, normal path delay, and path retransmission budget for other paths. For example, if the channel condition of a link is bad, the link weighting parameter can be increased to meet certain packet loss requirements. The enhanced a _{i, j} and b _{i, j} allow a greater number of retransmissions and a longer waiting time before the packet is discarded (within the limits allowed by the ARQ mechanism). As another example, the weighting parameters of links close to the base station or serving multiple relay stations and subscriber stations can be increased. Successful transmission over this link is important because this link is located at the root of the topology and thus affects many other links. As yet another example, the weighting parameters can be reduced to meet the overall delay requirements of the path. If the normal path delay and the path retransmission budget for other paths are known, those information can also be used. For example, if the sum of the normal path delay and the path retransmission budget for path m is greater than such total for path n, the path retransmission budget for path n can be increased by that difference.

本明細書に関して、パス遅延予算は、パケットが、基地局からアクセス局に至るパス内で伝送される間に、そのパケットによって消費され得る合計時間である。この時間のいくらかは、パケットの初期伝送によって消費される一方で、この時間の別のいくらかは、再送によって消費される。図３Ｂを再び参照すると、処理ロジックが、Ｔ_{ＴｘＯｆｆ}として表される伝送時間オフセットを算出する（処理ブロック３０４）。一実施形態において、処理ロジックは、以下のとおり、すべてのパスのパス遅延予算のなかで最大の値をとることによって、伝送時間オフセットを算出する。

ただし、Ｐは、ネットワーク上、又はトポロジ上のすべてのパスのセットである。 In one embodiment, path delay budget of each path, D _i is as follows, the normal path delay, and based on the path retransmission budget is calculated. That is,

For the purposes of this specification, the path delay budget is the total time that a packet can be consumed while being transmitted in the path from the base station to the access station. Some of this time is consumed by the initial transmission of the packet, while some of this time is consumed by retransmission. Referring back to FIG. 3B, processing logic calculates a transmission time offset, expressed as T _TxOff (processing block 304). In one embodiment, processing logic calculates the transmission time offset by taking the maximum value among the path delay budgets of all paths as follows.

Where P is a set of all paths on the network or topology.

  By taking the maximum path delay budget as a transmission time offset, at least an adaptation is made to the path retransmission budget for all paths. First, the path with the largest path delay budget is allocated a reserved time that can be used for retransmission that is exactly the same as that budgeted by the path retransmission budget. On the other hand, paths with path delay budgets less than the maximum are allocated more reserved time than was budgeted by the path retransmission budget. In order to take advantage of this large reserved time to favor further reliability, the path retransmission budget for the path, as well as other per-link delay parameters and per-path delay parameters, are included in optional processing block 305 and 306 and 306 and 306, respectively, processing logic adjusts per-path delay parameters if necessary, and processing logic adjusts per-link delay parameters if necessary. To do.

ただし、Ｐは、ネットワーク上、又はトポロジ上のすべてのパスのセットである。 In another embodiment, the transmission time offset is calculated by taking a value less than the maximum value among the path delay budgets of all paths.

Where P is a set of all paths on the network or topology.

  This means that, for example, while calculating the weighting parameters in processing block 303 of FIG. 3B, the overall delay requirement from the base station to the subscriber station is not taken into account, and the resulting path delay budget for that path is It can happen if it is larger than the overall delay requirement.

  For paths with a path delay budget greater than the transmission time offset, the path retransmission budget, as well as other per-path delay parameters and per-link delay parameters, are adapted to the adjusted path delay budget within the transmission time offset. It must be coordinated in optional processing blocks 305 and 306 to obtain. When these parameters are adjusted, it is ensured that the retransmission established by the path retransmission budget is handled within the transmission time offset. If no adjustment is made, there may be no response to a budgeted resend. For paths with a path delay budget smaller than the transmission time offset, in optional processing blocks 305 and 306, the path retransmission budget, and other per-path delay parameters and per-link delay parameters are adapted for further retransmissions. It can also be adjusted.

ただし、Ｐは、ネットワーク上、又はトポロジ上のすべてのパスのセットである。 The transmission time offset is usually not less than the normal path delay of any path so as to accommodate at least the time for the initial transmission of a packet over a link on the path.

Where P is a set of all paths on the network or topology.

  As described above, after calculating the transmission time offset, processing logic optionally adjusts the delay parameter for each path (processing block 305). As described with reference to processing block 304, this adjustment may increase or decrease the path retransmission budget for one or more paths.

に基づいて、調整される。例えば、調整されたパス遅延予算、

は、伝送時間オフセット及びリンクスケジューリングマージンに基づいて獲得され、調整されたパス再送予算、

は、以下のとおり、調整されたパス遅延予算と通常のパス遅延の差に基づいて、獲得される。すなわち、

処理ブロック３０６において、処理ロジックは、処理ブロック３０５を実行した結果に基づいて、リンクごとのパス遅延パラメータを調整する。一実施形態において、調整されたパス遅延予算、及び調整されたパス再送予算が、処理ブロック３０５において低減されるパスに関して、そのパス上のリンクのリンクごとの遅延パラメータは、再送のために消費される時間がより少ないことが可能であるように調整される。例えば、パスｉのリンクｊに関して、以下のリンクパラメータ、すなわち、リンク再送遅延、

、加重パラメータ、ａ_ｉ，ｊ、ｂ_ｉ，ｊ、ＡＲＱパラメータ、ＡＲＱ＿ＢＬＯＣＫ＿ＬＩＦＥＴＩＭＥ_ｉ，ｊ、ＲＥＴＲＹ＿ＬＩＭＩＴ_ｉ，ｊなどの１つ又は複数が、低減されることが可能である。別の実施形態において、調整されたパス遅延予算、及び調整されたパス再送予算が、処理ブロック３０５において増加されるパスに関して、そのパス上のリンクのリンクごとの遅延パラメータは、再送のために消費される時間がより多いことが可能であるように調整される。例えば、パスｉのリンクｊに関して、以下のリンクパラメータ、すなわち、リンク再送遅延

、加重パラメータ、ａ_ｉ，ｊ、ｂ_ｉ，ｊ、ＡＲＱパラメータ、ＡＲＱ＿ＢＬＯＣＫ＿ＬＩＦＥＴＩＭＥ_ｉ，ｊ、ＲＥＴＲＹ＿ＬＩＭＩＴ_ｉ，ｊなどの１つ又は複数が、増加されることが可能である。 In one embodiment, the path retransmission budget and path delay budget are optional link parameters that fine tune the delay parameters to account for transmission time offsets, normal path delays, and practical scheduling limits. Link scheduling margin for one link,

Adjusted based on For example, an adjusted path delay budget,

Is obtained and adjusted based on transmission time offset and link scheduling margin and adjusted path retransmission budget,

Is obtained based on the difference between the adjusted path delay budget and the normal path delay as follows. That is,

In processing block 306, processing logic adjusts the path delay parameter for each link based on the result of executing processing block 305. In one embodiment, for a path for which the adjusted path delay budget and the adjusted path retransmission budget are reduced in processing block 305, the per-link delay parameters of the links on that path are consumed for retransmission. Is adjusted so that less time is required. For example, for link j of path i, the following link parameters: link retransmission delay,

, Weighting parameters, a _{i, j} , b _{i, j} , ARQ parameters, ARQ_BLOCK_LIFETIME _{i, j} , RETRY_LIMIT _{i, j,} etc. can be reduced. In another embodiment, for a path for which the adjusted path delay budget and the adjusted path retransmission budget are increased in processing block 305, the per-link delay parameters of the links on that path are consumed for retransmission. It is adjusted so that more time can be played. For example, for link j of path i, the following link parameters: link retransmission delay

, Weighting parameters, a _{i, j} , b _{i, j} , ARQ parameters, ARQ_BLOCK_LIFETIME _{i, j} , RETRY_LIMIT _{i, j,} etc. can be increased.

  Note that the range of these adjustments is limited by the transmission time offset so that there is no increase or decrease in the total delay of the synchronized multilink transmission.

  FIG. 6 is a flow diagram of one embodiment of a process for handling a packet as it is received by a base station. This process is performed by processing logic that can comprise hardware (circuitry, dedicated logic, etc.), software (such as is run on a general purpose computer system or a dedicated machine), or a combination of hardware and software. Is done.

  Referring to FIG. 6, processing logic (eg, a receiver unit) at the base station receives the packet (processing block 601). Optionally, processing logic (eg, a packet classifier) classifies the packet to identify requirements associated with the packet (processing block 602). Such requirements may include, but are not limited to, QoS, ARQ, synchronized multilink transmission, routing, topology, power saving mode, etc., well known in the art. Not. The packet can be any type of PHY header, PHY trailer, MAC header, MAC trailer, IP header, IP trailer, TCP / UDP header, TCP / UDP trailer, or protocol unit in a manner well known in the art. It can be classified according to header, trailer or body values. In one embodiment, these requirements are associated with a set of sequential packets, including received packets, which are further referred to herein as connections, service flows, streams. The packet classification in this processing block can be connection, service flow, and stream identification.

  After packet classification, processing logic (eg, a scheduler) at the base station schedules the packet for transmission over one or more relay links determined by topology and routing (processing block 603). These relay links include all first links of all paths. In one embodiment, the scheduled relay transmission time of the first relay link on path i is in accordance with QoS requirements, adjusted or unadjusted path delay parameters, link scheduling margin, ARQ status, and ARQ window. To be determined. Depending on the topology, the packet can be scheduled to be transmitted multiple times over multiple relay links. For example, referring to the topology of FIG. 5, base station 501 is associated with two relay links, so packets are scheduled twice unless multicast or broadcast transmissions that cover both relay links are possible. Is done.

  In one embodiment, in a frame-based multi-hop network, packet scheduling includes determining a frame and, optionally, a position within the frame where the packet is to be transmitted. For example, in a frame-based OFDM network, such as 802.16, a packet may be scheduled to be transmitted in a frame with a frame number of 5, for example, and within that frame, the packet Can be transmitted using a set of subcarriers and a set of OFDM symbol times. Alternatively, if the network (or at least the base station and the relay station that is part of the topology) is synchronized to a common reference clock, the scheduled transmission time may be expressed in terms of the common reference clock. Is possible.

  In one embodiment, if a packet is scheduled multiple times, the packet is first transmitted over the path with the largest (unregulated) path delay budget and then the one with the smaller link scheduling margin. Are transmitted via other paths in order. Preferably, the difference between the transmission of the first transmission and the other path does not exceed the link scheduling margin associated with that path. This ensures that the transmission time offset calculated by FIG. 3B may be applicable, even if the packets are transmitted in different frames or at different reference times, depending on the path. .

  Referring again to FIG. 6, after scheduling transmission over the relay link, processing logic (eg, scheduler) at the base station determines whether synchronized multilink transmission is enabled for the packet. (Processing block 604). This operation is necessary in the network when some of the packets do not require synchronized multilink transmission. If all packets delivered in the network require synchronized transmission, this operation is not necessary and the scheduler can proceed directly from processing block 603 to processing block 605. The determination at processing block 604 may be straightforward depending on the classification of processing block 602. For example, at processing block 602, the packet may be classified as a connection with synchronized multilink transmission enabled. In such a case, synchronized multilink transmission is enabled for the packet.

ただし、Ｐは、ネットワーク上、又はトポロジ上のすべてのパスのセットである。 When synchronized multilink transmission is enabled for a packet, the process (and base station) is responsible for adjusting the scheduled relay transmission time, transmission time offset, other path or link adjustments at the base station. Based on the adjusted or unadjusted delay parameter, the synchronized transmission time, T _Stx, is calculated. In one embodiment, for example, on path i with the largest path delay budget, the scheduled relay transmission time of this path, eg, T _{ltx, i, l,} is used. For example,
T _Stx = T _{ltx, i, l} + T _TxOff
In another embodiment, the synchronized transmission time is calculated based on the scheduled relay transmission time and the adjusted path delay budget. For example,

Where P is a set of all paths on the network or topology.

  In one embodiment, transmission time offsets and other delay parameters are made available when or before processing block 605 is executed. For example, the transmission time offset can be calculated using the scheduling process of FIG. 3B when a connection associated with the packet is created. Alternatively, the transmission time offset may be calculated as it is processed in processing block 605 after the packet arrives.

  After the synchronized transmission time is calculated, at the scheduled relay transmission time, processing logic (eg, transmitter) at the base station transmits the packet and information at the synchronized transmission time (processing) Block 606). In one embodiment, the packet and information are transmitted together, for example, in the same frame, in the same burst, or in the same message. FIG. 7 shows an exemplary frame format in which packets and information are included in a single frame. Referring to FIG. 7, this frame format includes a header 711, a packet 702, a synchronized transmission time 703, an optional section 704 (eg, it may not be related to synchronized transmission, but the unused frame Storing other messages or other packets that can be placed to utilize space), and a trailer 705. If the packet is scheduled for transmission over multiple relay links, the packet may be transmitted multiple times depending on the scheduling decision made at processing block 603. If ARQ is enabled on the link on which the packet is transmitted, after transmission of the packet, one embodiment of the ARQ mechanism may be initiated and further performed in parallel with the packet processing of FIG. is there.

  FIG. 8 is a flow diagram of one embodiment of an exemplary ARQ mechanism for the transmitter. This process is performed by processing logic that can comprise hardware (circuitry, dedicated logic, etc.), software (such as is run on a general purpose computer system or a dedicated machine), or a combination of hardware and software. Is done.

  Referring to FIG. 8, processing logic transmits a packet, for example, in response to packet processing presented in the processing block of FIG. 6 (processing block 801). Processing logic may ensure that one or more ARQ state transitions occur in accordance with the ARQ described herein (processing block 802) and packet transmission is successful (and the process transitions to processing block 803). ) Or a packet transmission failure where the packet is discarded as part of processing block 804. Any ARQ scheme that does not consider synchronized multilink transmission is adapted to the ARQ described herein by adding the event that the ARQ state transitions to a state where pending packets are discarded. Is possible. In one embodiment, this event is defined as a situation where the packet is unlikely to reach the access station in time for the synchronized transmission time. In such an event, there is no merit in continuing the ARQ process or delivering packets to lower relay stations and access stations present on the forward path. Thus, the packet can be discarded to reduce unnecessary packet retransmissions. A new state transition captures such an observation, and if this event occurs, the packet is discarded at processing block 804. As long as this event does not occur, the ARQ state can transition according to normal ARQ (eg, according to the state diagram of FIG. 2), and the ARQ state transition can also occur according to normal ARQ to succeed in packet transmission. (Processing block 803), or packet transmission failure where the packet is discarded (processing block 804).

  When the packet is discarded at processing block 804, the base station sends an ARQ discard message for the packet, instructing the receiver to discard the packet (processing block 805). Upon receiving this message, the receiver discards the packet and, in one embodiment, sends an acknowledgment message in response to the ARQ discard message. Both the transmitter and the receiver can then send the ARQ window associated with that packet forward. This ensures that the two ARQ states (transmitter, eg, base station state and receiver state) are synchronized even if the packet is discarded.

  In FIG. 8, an exemplary event and associated state transitions according to one embodiment of the invention are shown. In one embodiment, this event is defined based on synchronized transmission time, current time, and parameters referred to herein as transmission thresholds. For example, this event occurs when the time remaining until the synchronized transmission time is less than the transmission threshold. This threshold may be: QoS requirements including link retransmission delay for links on the path, channel characteristics of links on the path, topology of multihop networks, delay requirements and / or packet loss requirements, normal path delay , And one or more of the path retransmission budgets for that path.

  In one embodiment, the transmission threshold is simply set to zero. In such cases, if the event occurs, that is, if the current time has passed the synchronized transmission time, the packet has already missed the synchronized transmission time. Thus, there is no need to continue the ARQ process or forward the packet.

  In another embodiment, the threshold is set to a normal path delay that is the sum of the normal link delays for links from the base station and access station. Consider the first path in the example topology of FIG. In such cases, when the event occurs, this means that the current time has not passed the synchronized transmission time, but the time remaining before the synchronized transmission time is the normal path. This means that the delay is not enough. Thus, there is no need to continue the ARQ process or forward the packet. In yet another embodiment, the threshold is set to the sum of the normal path delay and a reasonable time for retransmissions over one or more links in that path. In one embodiment, this reasonable time is a QoS requirement, including link retransmission delay for links on the path, channel characteristics of the links on the path, multi-hop network topology, delay requirements and / or packet loss requirements, Path delay, the path retransmission budget for the path, and the adjusted or unadjusted link or path delay parameters. In one embodiment, this reasonable time may be one or more link retransmission delays for one or more links in a path where channel errors are expected. As another example, this reasonable time can be the ARQ block lifetime for one or more links, where the channel status is very bad and the ARQ block lifetime can expire. is there.

FIG. 9 illustrates one embodiment of the exemplary ARQ state diagram of FIG. 2 modified to incorporate events and state transitions, according to one embodiment of the present invention. In this embodiment, the event is identified by the expiration of a timer whose value is obtained based on the relay transmission time T _Itx , the synchronized transmission time T _Stx , and the transmission threshold T _th . For example, after transmitting a packet at the relay transmission time T _Itx , the timer T _Timer is set to a value obtained by the following equation.
T _Timer = T _Stx −T _Itx −T _th
Because the exemplary ARQ state diagram of FIG. 2 has a timer for packet discard that expires after AQR_BLOCK_LIFETIME, the technique described herein uses the smaller of T _Timer and ARQ_BLOCK_LIFETIME for the packet discard. It can be incorporated by taking as a new timer value. Referring again to FIG. 6, if the base station determines that synchronized multilink transmission is not enabled for the packet, the process transitions to processing block 607 where processing logic at the base station (eg, , Transmitter) transmits the packet over the relay link at the scheduled relay transmission time calculated in process block 603. The packet may be transmitted once or multiple times according to the scheduling decision made at processing block 603. Also, if ARQ is enabled on the relay link to which the packet is transmitted, the ARQ mechanism is started after the packet is transmitted. Note that the ARQ mechanism may not be the ARQ described herein because synchronized multilink transmission is not enabled for the packet. For example, the exemplary ARQ of FIG. 2 may be used in processing block 607.

  After transmitting the packet, processing logic (eg, a base station scheduler) determines whether the base station serves as an access link for the packet (processing block 608). In the exemplary topologies of FIGS. 4 and 5, the base station does not serve as an access link. In such cases, the base station proceeds to processing block 611 and the processing of the packet ends. This determination may not need to be made for each packet, but is shown as such in FIG. For example, if this determination does not change throughout the operation of the base station or during the lifetime of the connection, the base station serves as the access link once for the entire operation or once for the lifetime of the connection. Can be determined and processing block 608 can be skipped at other times.

  If the base station acts as an access link for the packet, the process schedules transmission of the packet at the synchronized transmission time calculated by processing logic (e.g., base station scheduler). Transition to processing block 609. Next, processing logic (eg, a base station transmitter) transmits the packet over the access link as scheduled in processing block 609 (processing block 610). In one embodiment, the packet is sent simultaneously to all other access links (ie, synchronized links) in accordance with relay station packet processing as described below.

  FIG. 10 is a flow diagram of one embodiment of a process for processing packets received by a relay station. This process is performed by processing logic that can comprise hardware (circuitry, dedicated logic, etc.), software (such as is run on a general purpose computer system or a dedicated machine), or a combination of hardware and software. Is done.

  Referring to FIG. 10, processing logic at the relay station receives a packet (processing block 1001). The synchronized transmission time may also be available when the processing logic receives the packet if the base station is transmitting the synchronized transmission time with the packet. If the synchronized transmission time is transmitted separately, the synchronized transmission time must be available before the relay station reaches processing block 1006.

  After receiving the packet, optionally, processing logic classifies the packet to identify requirements associated with the packet (processing block 1002). As with processing block 602 of FIG. 6, these requirements can include QoS, ARQ, synchronized multilink transmission, routing, topology, and ARQ, and packet classification can , Service flow, and stream identification.

  After receiving the packet and performing any packet classification, processing logic at the relay station determines whether the relay station is the end of the path for all paths through the station (processing block 1003). For example, the relay station RS2 in the example topology of FIG. 5 is the end of the second path, but is not the end of the third path. On the other hand, RS3 (and also RS1) is the end of the path, with only one path going through the station. Although not shown in FIG. 10, this determination may not have to be made for each packet. For example, if this determination does not change throughout the operation of the relay station or during the lifetime of the connection, the relay station may execute processing block 1003 once for the entire operation or once for the lifetime of the connection. At other times, the processing block 1003 can be skipped.

  If the relay station is at the end of a path for all paths, no processing related to packet relaying (ie, processing blocks 1004-1009) is necessary, and therefore proceeds to processing block 1010, and if necessary, as an access link Process.

  If not, the process transitions to processing block 1004.

の部分を使用する。パスｉ上の第ｊ番の中継リンクのスケジュールされた中継伝送時間は、説明のため、本明細書ではＴ_{ｌｔｘ，ｉ，ｊ}と表される。 At processing block 1004, processing logic at the relay station schedules transmission of the packet over the next relay link as determined by topology, routing, etc. This process is basically the same as the process block 603 of FIG. 6, except that not the information about the base station but the information about the relay station is used. In one embodiment, the relay station in the j-th link of the i-th path does not use the base station topology and the link scheduling margin, but the topology related to the relay station, and the link scheduling margin,

Use the part. The scheduled relay transmission time for the jth relay link on path i is _denoted herein as T _{ltx, i, j} for purposes of illustration.

  After scheduling the transmission, processing logic at the relay station determines whether synchronized multilink transmission is enabled for the packet (processing block 1005). In one embodiment, this is performed in the same manner that the base station makes this determination in process block 604 of FIG.

If synchronized multilink transmission is enabled for the packet, the process transitions to processing block 1006 where processing logic at the relay station determines whether the packet is in time for the synchronized transmission time. . If the relay station is on multiple paths to multiple access stations, processing logic considers each path separately. In one embodiment, the relay station on the j-th link of the i-th path has a synchronized transmission time T _stx , a scheduled relay transmission time T _{ltx, i, j} , and a transmission threshold T _{th. , I, j} , it is determined that the packet is in time for path i. In one embodiment shown in FIG. 11, the relay station determines that if the synchronized transmission time T _stx is greater than the relay transmission time T _{ltx, i, j} by the transmission threshold T _{th, i, j} , the packet Judge that it is in time.

  This threshold may be: QoS requirements including: link retransmission delay for links on the forward path, channel characteristics of links on the forward path, multi-hop network topology, delay requirements and / or packet loss requirements; It can be determined based on one or more of the normal forward path delays and the path retransmission budget for that forward path.

  In one embodiment, the transmission threshold is simply set to zero. In such cases, if the relay transmission time is scheduled after the synchronized transmission time, the packet has already missed the synchronized transmission time.

  Therefore, it is determined that the packet is not in time.

  In another embodiment, this threshold is set to the normal forward path delay, which is the sum of the normal link delays for links from relay stations and access stations (ie, links in the forward path). Consider the third path in the example topology of FIG. In such cases, the threshold is set to the normal forward path delay of path 3 as seen from the relay station, which is the normal link delay of the second link of the third path. Unless the synchronized transmission time is greater than the relay transmission time by the normal link delay of the second link of the third path, the packet does not reach the access link (RS3) in time for the synchronized transmission. there is a possibility.

  In yet another embodiment, this threshold is set to the sum of the normal forward path delay and a reasonable time for retransmissions over one or more links in that forward path. This reasonable time includes link retransmission delay for the link, link channel characteristics, multi-hop network topology, QoS requirements including delay requirements and / or packet loss requirements, normal forward path delay, and path retransmission budget for that path. , And the adjusted or unadjusted link or path delay parameters. With respect to the example topology of FIG. 5, in one embodiment, this reasonable time may be one or more link retransmission delays over the second link of the third path.

  If there is a path that is a subpath of another path, processing block 1005 may be performed only for that subpath. This is because the sub-path is the shorter of the two paths, and if a packet is in time for one path, the packet is in time for whether the packet is in time for the other path. Regardless, it must be transmitted over the next relay link. If the two paths do not share any relay links, processing block 1005 can be performed independently for each of the two paths. Depending on the result of processing block 1005, subsequent processing blocks, processing block 1007 or 1008, may be executed independently. For example, a packet may be in time for one path but not in time for another path. In such cases, the packet is transmitted only over the timely path.

  If the packet is in time for a path, processing logic at the relay station transmits the packet and information over the relay link at a synchronized transmission time (processing block 1008). As with processing block 606 of FIG. 6, the packets and information can be transmitted together, for example, in the same frame, in the same burst, or in the same message. If a packet is scheduled to be transmitted over multiple relay links, the packet may be transmitted multiple times according to the scheduling decision made at processing block 1004. If ARQ is enabled on the link on which the packet is transmitted, after transmission of the packet, the ARQ mechanism described herein is initiated and further performed in parallel with subsequent packet processing. It is possible. Any ARQ of the embodiment of the invention shown in connection with FIG. 6 may be used by the relay station. After transmission, the relay station proceeds to processing block 1010.

  If the packet is not in time for the path, processing logic at the relay station cancels the scheduled packet transmission over the path (processing block 1007) and proceeds to processing block 1013. Since the packet is unlikely to reach the access link by the synchronized transmission time, it is not necessary to transmit the packet. Again, this is another aspect of the techniques described herein that reduce unnecessary packet transmissions.

  If synchronized multilink transmission is not enabled for the packet, the process proceeds to processing block where processing logic transmits the packet over the relay link at the relay transmission time scheduled in processing block 1004. Transition to 1009. The packet can be transmitted one or more times according to scheduling decisions made in the processing block. Also, if ARQ is enabled on the relay link on which the packet is transmitted, the ARQ mechanism is started after transmission of the packet, but synchronized multilink transmission is enabled for the packet. Therefore, this ARQ mechanism may not be the ARQ mechanism described above.

  Next, processing logic determines whether the relay station acts as an access link for the packet (processing block 1010). If the relay station does not act as an access link, the process transitions to processing block 1013 and the processing of the packet ends.

  If the relay station acts as an access link for the packet, the process transitions to processing block 1011 where processing logic schedules transmission of the packet at the synchronized transmission time received from the base station. . Next, processing logic transmits the packet over the access link scheduled in processing block 1011 (processing block 1012). Note that the base station, and all other access stations, transmit their packets simultaneously over their synchronized links.

  FIG. 12 is a block diagram of one embodiment of a station. In one embodiment, the station is a base station that operates as described in FIG. 6 above. Referring to FIG. 12, the base station 1200 includes a receiver 1201, a packet classifier 1202, a scheduler 1203, and a transmitter 1204. The receiver 1201 receives the packet. Packet classifier 1202 classifies the packet to identify requirements associated with the packet. This is optional. These packets are stored in the buffer memory 1210. Next, scheduler 1203 schedules the packet to be transmitted over one or more relay links and / or access links as described above in connection with FIG. After the transmission time is calculated, at the scheduled transmission time, the transmitter 1204 transmits the packet and / or the information at the synchronized transmission time via their relay link and / or access link. To do. Note that a control unit (not shown) controls each of these units. Note that the station is a relay station and can further operate as described above in FIG.

  While alternatives and modifications of the invention will no doubt become apparent to those skilled in the art after reading the foregoing description, any particular embodiment shown and described by way of example is intended to be limiting. It should be understood that it is not intended at all. Therefore, the details of the various embodiments are not intended to limit the scope of the claims, which are only described as essential to the invention.

Claims (2)

A method used in an ARQ (automatic repeat request) compliant multi-hop network that supports synchronized transmission over a specific set of synchronized links, comprising:
Performing pre-transmission of packets to a plurality of hops, allowing the plurality of hops in a base station and the network to transmit packets synchronously to one or more mobile stations in a wireless communication system;
In one or more sets of hops forming a path in the network, if the remaining time until the synchronized transmission time for the entire path is greater than a threshold, retransmit the packet one or more times Steps to perform;
Including methods. An article of manufacture having one or more computer-readable storage media storing a plurality of instructions,
The plurality of instructions cause the system to execute a method used in an ARQ (automatic repeat request) enabled multi-hop network that supports synchronized transmission over a specific set of synchronized links;
The method
Performing pre-transmission of packets to a plurality of hops, allowing the plurality of hops in a base station and the network to transmit packets synchronously to one or more mobile stations in a wireless communication system;
In one or more sets of hops forming a path in the network, if the remaining time until the synchronized transmission time for the entire path is greater than a threshold, retransmit the packet one or more times Steps to perform;
including,
A manufactured product characterized by that.

Images